UC Berkeley UC Berkeley Previously Published Works

Title Imaging the Centromedian Thalamic Nucleus Using Quantitative Susceptibility Mapping.

Permalink https://escholarship.org/uc/item/7pm0m9bk

Authors Li, Jun Li, Yufei Gutierrez, Lorenzo et al.

Publication Date 2019

DOI 10.3389/fnhum.2019.00447

Peer reviewed

eScholarship.org Powered by the California Digital Library University of California BRIEF RESEARCH REPORT published: 09 January 2020 doi: 10.3389/fnhum.2019.00447

Imaging the Centromedian Thalamic Nucleus Using Quantitative Susceptibility Mapping

Jun Li 1†, Yufei Li 2†, Lorenzo Gutierrez 2, Wenying Xu 1, Yiwen Wu 3, Chunlei Liu 4,5, Dianyou Li 1, Bomin Sun 1, Chencheng Zhang 1* and Hongjiang Wei 2*

1Department of Functional Neurosurgery, Ruijin Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China, 2Institute for Medical Imaging Technology, School of Biomedical Engineering, Shanghai Jiao Tong University, Shanghai, China, 3Department of Neurology, Ruijin Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, Edited by: China, 4Department of Electrical Engineering and Computer Sciences, University of California, Berkeley, Berkeley, CA, United 5 Adolfo Ramirez-Zamora, States, Helen Wills Neuroscience Institute, University of California, Berkeley, Berkeley, CA, United States University of Florida Health, United States The centromedian (CM) nucleus is an intralaminar thalamic nucleus that is considered as Reviewed by: a potentially effective target of deep brain stimulation (DBS) and ablative surgeries for the Wolf-Julian Neumann, Charité Medical University of Berlin, treatment of multiple neurological and psychiatric disorders. However, the structure of Germany CM is invisible on the standard T1- and T2-weighted (T1w and T2w) magnetic resonance Zhijiang Wang, Peking University Sixth Hospital, images, which hamper it as a direct DBS target for clinical applications. The purpose of China the current study is to demonstrate the use of quantitative susceptibility mapping (QSM) *Correspondence: technique to image the CM within the thalamic region. Twelve patients with Parkinson’s Chencheng Zhang disease, dystonia, or schizophrenia were included in this study. A 3D multi-echo gradient [email protected] Hongjiang Wei recalled echo (GRE) sequence was acquired together with T1w and T2w images on a [email protected] 3-T MR scanner. The QSM image was reconstructed from the GRE phase data. Direct †These authors have contributed visual inspection of the CM was made on T1w, T2w, and QSM images. Furthermore, equally to this work the contrast-to-noise ratios (CNRs) of the CM to the adjacent posterior part of Specialty section: on T1w, T2w, and QSM images were compared using the one-way analysis of variance This article was submitted to Brain Imaging and Stimulation, a section of (ANOVA) test. QSM dramatically improved the visualization of the CM nucleus. Clear the journal Frontiers in Human delineation of CM compared to the surroundings was observed on QSM but not on T1w Neuroscience and T2w images. Statistical analysis showed that the CNR on QSM was significantly

Received: 17 September 2019 higher than those on T1w and T2w images. Taken together, our results indicate that Accepted: 05 December 2019 QSM is a promising technique for improving the visualization of CM as a direct targeting Published: 09 January 2020 for DBS surgery. Citation: Li J, Li Y, Gutierrez L, Xu W, Wu Y, Keywords: deep brain stimulation, direct targeting, gradient recalled echo, quantitative susceptibility mapping, Liu C, Li D, Sun B, Zhang C and Wei centromedian nucleus H (2020) Imaging the Centromedian Thalamic Nucleus Using Quantitative Susceptibility Mapping. Abbreviations: CM, centromedian nucleus; CNR, contrast-to-noise ratio; DBS, deep brain stimulation; GRE, gradient Front. Hum. Neurosci. 13:447. recalled echo; MRI, magnetic resonance imaging; QSM, quantitative susceptibility mapping; T1w, T1-weighted; T2w, doi: 10.3389/fnhum.2019.00447 T2-weighted.

Frontiers in Human Neuroscience| www.frontiersin.org 1 January 2020 | Volume 13 | Article 447 Li et al. Imaging the Centromedian Nucleus

INTRODUCTION accurate structural delineation (Liu et al., 2015). QSM has been clinically used to assess important tissue functions and disease The centromedian nucleus (CM) or centromedian–parafasicular (Wang et al., 2017), and recently it has been demonstrated nucleus complex, located in the caudal intralaminar thalamic for improving the depiction of DBS target structures with nuclei, has been reported to be a potentially effective target iron-rich nucleus (paramagnetic), e.g., the for deep brain stimulation (DBS) or ablative surgeries for the (Liu et al., 2013; Alkemade et al., 2017) and the treatment of various neurological and psychiatric diseases, internus (Wei et al., 2019), with the surrounding white matters e.g., Parkinson’s disease, Tourette syndrome, generalized (diamagnetic). The thalamus contains different subregions that epilepsy, and intractable neuropathic pain (Ilyas et al., 2019). are known to have various iron deposits and different degrees However, the surgeries targeting CM still relied on the indirect of myelinated white matters (Morris et al., 1992; Zhang et al., targeting method by registering a normalized atlas to the patient’s 2018), which indicates that QSM, by using the susceptibility magnetic resonance imaging (MRI) data and then the CM differences existing between substructures, may be a proper coordinates are used for target localization (Krauss et al., 2002; imaging technique to identify CM. Kim et al., 2017; Sharma et al., 2017). This indirect targeting The aim of this study is to examine whether QSM could method may lead to suboptimal targeting since significant delineate the CM nucleus from its adjacent thalamic structures variations exist in brain structures between patients, and this and thus generate a direct visualization of the CM. variation causes unpredictable registration errors (Kennedy et al., 1998) and may sub-optimize treatment effect and increase MATERIALS AND METHODS the rate of surgical complications and adverse side effects (Chan et al., 2009). Human Subjects Direct targeting can improve the targeting accuracy in certain Twelve patients (six males and six females, mean age aspects as revealed by some studies (Tonge et al., 2016; Fenoy 41.8 ± 21.2 years old) with Parkinson’s disease (n = 5, mean and Schiess, 2018). Direct targeting requires that the anatomical age 61.0 ± 16.6), dystonia (n = 4, mean age 32.8 ± 8.6), or locations can be visible on certain image contrast. However, schizophrenia (n = 3, mean age 21.7 ± 10.3) were included as direct visualization of the CM nucleus using the standard T1w convenient samples in this study. Demographic information and T2w MRI sequences is challenging. On one hand, the volume collection and neuroradiological investigation were performed of the CM is small (smaller than 10 mm in most dimensions; by specialized movement disorder neurologists or psychiatrists. Ilyas et al., 2019). On the other hand, the contrast between The study was approved by the ethics committee of Ruijin the CM nucleus and its surrounding thalamic structures is Hospital, School of Medicine, Shanghai Jiao Tong University. pretty low. The absence of an imaging technique for direct All subjects provided written consent in accordance with the visualization of CM hampers the targeting accuracy of CM for Declaration of Helsinki. DBS surgery. Some researchers have made considerable efforts to improve Data Acquisition the individualized depiction of thalamic substructures. Lemaire Imaging was performed on a 3.0-T MR scanner equipped with et al.(2010) reported that high-resolution T1w images could be a 24-channel head coil. Each subject lay supine with their head used to image the substructures of the thalamus, which were very snugly fixed with foam pads. The subject was asked to keep still as comparable to myelin-stained histologic sections. However, the long as possible. 3D T1w and axial T2w images were acquired. A scan time for the protocol was approximately 14 h, which is not multi-echo GRE sequence was also performed. Detailed imaging suitable for routine clinical scans. Kanowski et al.(2010) showed parameters, including the time of repetition, time of echo, field that the CM is identifiable in a reasonable measurement time of view, voxel size, and total duration of scanning for the three of 13–26 min with two-dimensional high-resolution proton- imaging modalities, are summarized in Table 1. attenuation-weighted images at 3 T. However, only a few slices in axial plane covering the localized areas were acquired, which Image Processing still challenges targeting localization when using the surgical QSM images were reconstructed from GRE phase data. The planning software involving the 3D image registration procedure. details of QSM processing has been documented in the previous Bender et al.(2011) demonstrated that the CM could be roughly articles (Wei et al., 2015, 2017). In brief, three major steps identified by optimized 3D MPRAGE protocol, which would take were taken for the reconstruction of the QSM image. First, about 20 min to be acquired; however, clear discrimination of all the phase images of GRE were unwrapped using a Laplacian- thalamic substructures were not achievable. If anatomic imaging- based phase unwrapping. Afterward, the magnitude images were based targeting methods can be further improved, the accuracy used to extract the brain tissue using the FMRIB Software and efficiency of target selection for DBS or ablative surgeries Library Brain Extraction Tool1. Then, the background phases may further increase. were removed using the V_SHARP method to obtain the local Quantitative susceptibility mapping (QSM) reconstructed tissue phase images (Li et al., 2015). Finally, susceptibility maps from the MRI phase images of the 3D gradient recalled echo were generated after dipole inversion using streaking artifact (GRE) sequences could improve tissue contrast compared to T2w reduction for QSM method (STAR-QSM; Wei et al., 2015). images. QSM employed deconvolution of GRE phase images and removed the non-local susceptibility effects, depicting more 1https://fsl.fmrib.ox.ac.uk/fsl/fslwiki/BET

Frontiers in Human Neuroscience| www.frontiersin.org 2 January 2020 | Volume 13 | Article 447 Li et al. Imaging the Centromedian Nucleus

TABLE 1 | Imaging parameters. Parameter 3D T1w 2D T2w 3D GRE Imaging plane Axial Axial Axial Field of vision (mm) 240 240 240 240 240 240 × × × Matrix 320 320 320 320 320 320 × × × Resolution (mm) 0.75 0.75 1.5 0.75 0.75 1.5 0.75 0.75 1.5 × × × × × × Time of repetition (ms) 7.04 3,000/4,000 32.80 Time of echo (ms) 3.47 128.60/106.03 11.00 Scan time (s) 172 346 528

Image Inspection and Data Analysis posterior parts of the thalamus are also visible owing to different Firstly, we compared the QSM images to a schematic drawing magnetic susceptibility values, as delineated in Figure 1C. referenced from the overlay of Schaltenbrand and Wahren Figure 2 compares the contrast of CM on T1w, T2w, and histologic atlas (Schaltenbrand et al., 1977) to confirm whether QSM images at one representative section of a representative the CM can be visible on the QSM image. To calculate the patient. The location of CM nucleus is difficult to be identified contrast-to-noise ratios (CNRs), QSM and T2w images were on the T1w or T2w images. However, QSM image clearly firstly registered to the T1w image. Then, the regions of CM and shows the substructures of the thalamus, for example, the adjacent posterior thalamic tissues were manually defined medial, lateral, and posterior parts of the thalamus. The as masks on the QSM image (Supplementary Figure S1). CM nucleus is delineable from its surroundings on the Afterward, the masks of CM and posterior thalamus were QSM image. Clear delineation of CM and the surrounding applied to the T1w and T2w images. The CNRs of the CM tissues is attributed to the susceptibility difference existed nucleus referenced to the posterior thalamus were measured: between iron-rich nucleus and the adjacent myelinated CNR = |SCM−SpTH|/σ, where SCM and SpTH, respectively, white fiber . The QSM image exhibits a diamagnetic represent the mean signal intensities of the CM nucleus and susceptibility within the CM and a relatively paramagnetic posterior part of thalamus. σ represents noise measurement susceptibility of its surrounding thalamic tissues. The calculated as the standard deviation of the signal intensities in T1w, T2w, and QSM images at one representative section the posterior part of thalamus. The volumes of the CM nucleus containing CM nucleus on each patient are presented in the were also calculated on QSM images, by multiplying the number Supplementary Figure S2. of CM voxels and the voxel size.

Statistical Analysis A one-way analysis of variance (ANOVA) was used to compare the difference in CNRs among the three MR image modalities (IBM SPSS Statistics, version 22). If the one-way ANOVA gave a significant result, independent two-sample t-tests were further used as the post hoc tests to reveal the CNR differences between each two modalities (T1w vs. T2w, T1w vs. QSM, and T2w vs. QSM). Two-way repeated-measure ANOVAs were also performed to examine the significance of interaction between image modality (T1w, T2w, and QSM) and patient type (Parkinson’s disease, dystonia, and schizophrenia), and the significance of interaction between CM volume (left CM and right CM volumes) and patient type. The threshold of significance was set at p < 0.05.

RESULTS FIGURE 1 | The visualization of CM within the thalamus on quantitative Figure 1A shows a schematic drawing of thalamus that susceptibility mapping (QSM) image. (A) A schematic drawing of the CM and is referenced to the Schaltenbrand and Wahren atlas its surrounding thalamic structures, referenced to the overlay of the (Schaltenbrand et al., 1977). The diamagnetic CM is surrounded Schaltenbrand and Wahren atlas (Schaltenbrand et al., 1977). (B) An axial view of a slice of QSM image with thalamic substructures on a representative by the relatively paramagnetic medial, lateral, and posterior parts patient. (C) Enlarged view of thalamic substructures with the anatomical of the thalamus (Figure 1A). Figures 1B,D show the QSM image boundaries of CM and its surrounding thalamic parts (medial, lateral, and of one representative patient. As shown, the QSM image provides posterior) delineated. (D) Enlarged view of thalamic substructures. The a clear visualization on the anatomical structure of the CM (as anatomical location of CM nucleus is pointed by an orange arrow. indicated by an orange arrow) in a patient with Parkinson’s Abbreviations: CM, centromedian nucleus. L, lateral part of thalamus; M, medial part of thalamus; P, posterior part of thalamus. disease. The anatomical boundaries of the medial, lateral, and

Frontiers in Human Neuroscience| www.frontiersin.org 3 January 2020 | Volume 13 | Article 447 Li et al. Imaging the Centromedian Nucleus

FIGURE 2 | Comparison of the visualization of the CM nucleus on T1w, T2w, and QSM images. Axial slice views (upper row) and enlarged views of the thalamus (lower row) on T1w, T2w, and QSM images at one representative section on a representative patient. Abbreviations: CM, centromedian nucleus; QSM, quantitative susceptibility mapping.

The CNRs of the CM nucleus to the posterior part of thalamus are 0.37 ± 0.35, 0.67 ± 0.43, and 3.43 ± 0.49, respectively, on T1w, T2w, and QSM images (Figure 3). The ANOVA reveals significant differences among T1w, T2w, and QSM images in terms of the CNR, F(2) = 177.14, p < 0.001 (Figure 3). Post hoc tests (independent two-sample t-tests) indicate significant different CNRs between QSM and T1w (t(11) = 16.66, p < 0.001), and between QSM and T2w (t(11) = 17.44, p < 0.001). The mean CNRs for each type of patients are illustrated in the Supplementary Table S1, in which increased CNRs on QSM images are indicated in each of the three patient types. The mean volumes of the left and right CM nuclei are 160.95 ± 29.98 mm3 and 169.73 ± 50.34 mm3, respectively, as detected on the QSM images. No significant main effects of patient type on CNR value or CM volume, or interactions between patient type and CNR value, or between patient type and CM volume were found in our sample (ps > 0.142, Supplementary Tables S1, S2).

DISCUSSION FIGURE 3 | The CNRs of the CM to the posterior part of thalamus on the T1w, T2w, and QSM images. The dots represent the individual values of the The results demonstrate that with the QSM technique, Parkinson’s disease patients (square dots), the dystonia patients (circular the CM can be clearly delineated from the surrounding dots), and the schizophrenia patients (triangle dots). ∗∗∗Indicates p < 0.001. Abbreviations: CM, centromedian nucleus; pTH, posterior thalamus; QSM, subthalamic nuclei. Compared with commonly used T1w and quantitative susceptibility mapping. T2w images for DBS planning, QSM significantly improved the CNR of CM nucleus compared to its surrounding thalamic structures, suggesting that a QSM-based image is Aside from the surgical targets routinely used in clinical more suitable to target the patient-specific CM in DBS treatment (e.g., subthalamic nucleus, nucleus accumbens), there surgery directly. are some other targets with potential effectiveness in treating

Frontiers in Human Neuroscience| www.frontiersin.org 4 January 2020 | Volume 13 | Article 447 Li et al. Imaging the Centromedian Nucleus neurological and psychiatric diseases. The CM nucleus or of QSM imaging in clinical settings for relevant diseases centromedian–parafasicular nucleus complex, situated within should be given consideration by radiologists, neurosurgeons, the intralaminar nuclei of the thalamus, has abundant fiber MR manufacturers, and engineers. On the other hand, the connections with other thalamic nuclei, , and QSM technique has plenty of room to improve on for cerebral cortex (Ilyas et al., 2019). In several studies, the CM clinical applications, including shortening acquisition time and nucleus has been suggested as a potentially effective DBS reducing streaking artifacts to further improve the image quality target for the treatment of Parkinson’s disease (Caparros- (Wang et al., 2017). Lefebvre et al., 1999; Mazzone et al., 2006; Peppe et al., The signal intensity on a QSM image depends on the tissue 2008; Stefani et al., 2009) and Tourette syndrome (Houeto magnetic susceptibility (Wang and Liu, 2015). Due to the rich et al., 2005; Savica et al., 2012; Testini et al., 2016; Marano abundancy of iron in the blood, the blood vessel on a QSM et al., 2019). The clinical surgeries targeting at CM also image has a much higher intensity than gray matter, white show treatment effect for the generalized epilepsy (Fisher matter, or cerebrospinal fluid (Haacke et al., 2015). The visual et al., 1992; Velasco et al., 2007; Valentín et al., 2013; Li identification of the CM nucleus in the present sample is and Cook, 2018) and intractable neuropathic pain (Young unaffected by the blood vessels nearby. Furthermore, strong et al., 1995; Hollingworth et al., 2017) by means of DBS QSM signal can be observed in the structures with bleeding or thalamotomy. The DBS surgery targeting the CM nucleus or vascular dysmorphia (Liu et al., 2012, 2015; Chen et al., currently uses indirect ways in which a two-dimensional 2014). Although not being observed in the individuals of our stereotactic atlas of the thalamus is superimposed on a CT or sample, the delineation of the thalamic structures, including MRI scan relative to coarse anatomical landmarks including the CM nucleus, could be blurred in individuals with micro- anterior and posterior commissures (Stefani et al., 2009; Son bleeding or vascular dysmorphia at or around the regions et al., 2016; Testini et al., 2016). The indirect method of of interest. targeting the CM nucleus is due to the small volume of There are some limitations in the current study. The 3D the CM, measuring smaller than 10 mm in most dimensions GRE sequences is quite sensitive to patients’ motion during (Ilyas et al., 2019), and low image contrast between the CM the scan, and thus the application in patients with obvious nucleus and its surrounding thalamic structures on conventional tremor might be limited. The next limitation is that the scanning MRI images. The challenge of precisely locating the nucleus process for whole-brain QSM takes nearly 5–10 min. Although would limit the clinical application and the efficacy of CM- it is faster than the other methods that can also demonstrate DBS. Inter-patient variability may affect the accuracy of the the CM nucleus (Kanowski et al., 2010; Lemaire et al., 2010; placement DBS electrodes, and may sub-optimize treatment Bender et al., 2011), more rapid QSM techniques are yet to effect and increase the rate of surgical complications and be invented for DBS targeting in clinical application (Wei adverse side effects (Chan et al., 2009). Direct imaging CM et al., 2019). Another limitation is that the segmentations can be of great help for direct targeting of this intralaminar were done manually in this study. In future studies, the QSM thalamic nucleus. images could be normalized to MNI space and segmented Recently developed QSM image reconstructed from the based on available subcortical 3D atlases, e.g., using Lead-DBS GRE-sequence image is an effective technique that takes toolbox2. Finally, the sample size of the present study is advantage of differentiated iron concentration in different relatively small. However, even with small sample size, the subcortical microstructures to identify their locations (Deistung superiority of QSM for depicting CM nucleus can still be et al., 2017). Thalamic nuclei have sufficient iron concentration observed. The negative results of CNR values and CM volumes and different nuclei are with different levels of iron deposits between different types of patients may be attributed to the (Drayer et al., 1986; Morris et al., 1992). Thus, QSM can delineate limited sample size. Future studies with large sample sizes one nucleus from its adjacent myelinated axons, are needed to reveal the profiles of CNR values and CM such as for imaging the CM in this study. The delineation volumes in different types of patients, particularly in the patients of CM is attributed to the susceptibility difference existed in where DBS has shown potential effectiveness (e.g., Parkinson’s iron concentration compared to the adjacent myelin sheath disease, Tourette syndrome, generalized epilepsy, and intractable fibers. Although CM nucleus is also visible on high-resolution neuropathic pain). T1w images, 2-D proton-attenuation-weighted images, or images acquired by optimized 3D MPRAGE protocol (Kanowski et al., CONCLUSION 2010; Lemaire et al., 2010; Bender et al., 2011), those images usually would take at least 20 min (or even hours) to be acquired. In summary, we have demonstrated that the QSM images GRE image of the whole brain can be acquired within less than provide a significantly clearer visualization of the CM nucleus 10 min, which is more realistic for routine clinical scans for than T1w and T2w images, suggesting that a QSM image DBS planning. is likely more suitable to aid directly determining patient- Based on our finding that QSM could provide direct specific CM coordinates in the DBS and ablative surgeries. visualization on CM nucleus, together with the recent findings Future studies are highly needed to evaluate the QSM imaging that QSM could also provide superior anatomical delineation in CM nucleus on a large sample size, particularly in the subthalamic nucleus (Liu et al., 2013; Alkemade et al., 2017) and globus pallidus internus (Wei et al., 2019), the implementation 2www.lead-dbs.org

Frontiers in Human Neuroscience| www.frontiersin.org 5 January 2020 | Volume 13 | Article 447 Li et al. Imaging the Centromedian Nucleus types of patients who might potentially benefit from DBS AUTHOR CONTRIBUTIONS treatment, and confirm whether QSM technique can improve DBS targeting accuracy or effectiveness compared with indirect YW, DL, BS, CZ, and HW conceived and designed the study. WX targeting methods. and HW collected the data. JL, YL, LG, WX, and CL analyzed the data. JL, YL, LG, YW, CL, DL, BS, CZ, and HW interpreted the data and wrote the article. DATA AVAILABILITY STATEMENT

The datasets generated for this study are available on request to FUNDING the corresponding author. This work was supported by the Natural Science Foundation of China Grant (grant number 61901256) and SJTU Trans-med ETHICS STATEMENT Award Research.

The studies involving human participants were reviewed SUPPLEMENTARY MATERIAL and approved by the ethics committee of Ruijin Hospital, School of Medicine, Shanghai Jiao Tong University. The The Supplementary Material for this article can be found patients/participants provided their written informed consent to online at: https://www.frontiersin.org/articles/10.3389/fnhum. participate in this study. 2019.00447/full#supplementary-material.

REFERENCES Ilyas, A., Pizarro, D., Romeo, A. K., Riley, K. O., and Pati, S. (2019). The centromedian nucleus: , physiology and clinical implications. J. Clin. Alkemade, A., de Hollander, G., Keuken, M. C., Schafer, A., Ott, D. V. M., Neurosci. 63, 1–7. doi: 10.1016/j.jocn.2019.01.050 Schwarz, J., et al. (2017). Comparison of T2∗-weighted and QSM contrasts Kanowski, M., Voges, J., and Tempelmann, C. (2010). Delineation of the nucleus in Parkinson’s disease to visualize the STN with MRI. PLoS One 12:e0176130. centre median by proton density weighted magnetic resonance imaging at doi: 10.1371/journal.pone.0176130 3 T. Oper. Neurosurg. 66:ons-E121-ons-E123. doi: 10.1227/01.neu.0000348560. Bender, B., Manz, C., Korn, A., Nagele, T., and Klose, U. (2011). Optimized 85056.63 3D magnetization-prepared rapid acquisition of gradient echo: identification Kennedy, D. N., Lange, N., Makris, N., Bates, J., Meyer, J., and Caviness, V. S.Jr. of thalamus substructures at 3T. AJNR Am. J. Neuroradiol. 32, 2110–2115. (1998). Gyri of the human neocortex: an MRI-based analysis of volume and doi: 10.3174/ajnr.a2705 variance. Cereb. Cortex 8, 372–384. doi: 10.1093/cercor/8.4.372 Caparros-Lefebvre, D., Blond, S., Feltin, M. P., Pollak, P., and Benabid, A. L. Kim, S. H., Lim, S. C., Yang, D. W., Cho, J. H., Son, B. C., Kim, J., (1999). Improvement of levodopa induced dyskinesias by thalamic deep brain et al. (2017). Thalamo-cortical network underlying deep brain stimulation of stimulation is related to slight variation in electrode placement: possible centromedian thalamic nuclei in intractable epilepsy: a multimodal imaging involvement of the centre median and parafascicularis complex. J. Neurol. analysis. Neuropsychiatr. Dis. Treat. 13, 2607–2619. doi: 10.2147/ndt.s148617 Neurosurg. Psychiatry 67, 308–314. doi: 10.1136/jnnp.67.3.308 Krauss, J. K., Pohle, T., Weigel, R., and Burgunder, J. M. (2002). Deep Chan, D. T., Zhu, X. L., Yeung, J. H., Mok, V. C., Wong, E., Lau, C., et al. (2009). brain stimulation of the centre median-parafascicular complex in patients Complications of deep brain stimulation: a collective review. Asian J. Surg. 32, with movement disorders. J. Neurol. Neurosurg. Psychiatry 72, 546–548. 258–263. doi: 10.1016/S1015-9584(09)60404-8 doi: 10.1136/jnnp.72.4.546 Chen, W., Zhu, W., Kovanlikaya, I., Kovanlikaya, A., Liu, T., Wang, S., et al. (2014). Lemaire, J. J., Sakka, L., Ouchchane, L., Caire, F., Gabrillargues, J., and Bonny, J. M. Intracranial calcifications and hemorrhages: characterization with quantitative (2010). Anatomy of the human thalamus based on spontaneous contrast and susceptibility mapping. Radiology 270, 496–505. doi: 10.1148/radiol.13122640 microscopic voxels in high-field magnetic resonance imaging. Neurosurgery 66, Deistung, A., Schweser, F., and Reichenbach, J. R. (2017). Overview of quantitative 161–172. doi: 10.1227/01.neu.0000365617.41061.a3 susceptibility mapping. NMR Biomed. 30:e3569. doi: 10.1002/nbm.3569 Li, M. C. H., and Cook, M. J. (2018). Deep brain stimulation for drug-resistant Drayer, B., Burger, P., Darwin, R., Riederer, S., Herfkens, R., and Johnson, G. A. epilepsy. Epilepsia 59, 273–290. doi: 10.1111/epi.13964 (1986). MRI of brain iron. Am. J. Roentgenol. 147, 103–110. doi: 10.2214/ajr. Li, W., Wang, N., Yu, F., Han, H., Cao, W., Romero, R., et al. (2015). A method 147.1.103 for estimating and removing streaking artifacts in quantitative susceptibility Fenoy, A. J., and Schiess, M. C. (2018). Comparison of tractography-assisted mapping. Neuroimage 108, 111–122. doi: 10.1016/j.neuroimage.2014.12.043 to atlas-based targeting for deep brain stimulation in essential tremor. Mov. Liu, C., Wei, H., Gong, N. J., Cronin, M., Dibb, R., and Decker, K. Disord. 33, 1895–1901. doi: 10.1002/mds.27463 (2015). Quantitative susceptibility mapping: contrast mechanisms and clinical Fisher, R. S., Uematsu, S., Krauss, G. L., Cysyk, B. J., McPherson, R., applications. Tomography 1, 3–17. doi: 10.18383/j.tom.2015.00136 Lesser, R. P., et al. (1992). Placebo-controlled pilot study of centromedian Liu, T., Surapaneni, K., Lou, M., Cheng, L., Spincemaille, P., and Wang, Y. (2012). thalamic stimulation in treatment of intractable seizures. Epilepsia 33, 841–851. Cerebral microbleeds: burden assessment by using quantitative susceptibility doi: 10.1111/j.1528-1157.1992.tb02192.x mapping. Radiology 262, 269–278. doi: 10.1148/radiol.11110251 Haacke, E. M., Liu, S., Buch, S., Zheng, W., Wu, D., and Ye, Y. (2015). Quantitative Liu, T., Eskreis-Winkler, S., Schweitzer, A. D., Chen, W., Kaplitt, M. G., susceptibility mapping: current status and future directions. Magn. Reson. Tsiouris, A. J., et al. (2013). Improved subthalamic nucleus depiction Imaging 33, 1–25. doi: 10.1016/j.mri.2014.09.004 with quantitative susceptibility mapping. Radiology 269, 216–223. Hollingworth, M., Sims-Williams, H. P., Pickering, A. E., Barua, N., and doi: 10.1148/radiol.13121991 Patel, N. K. (2017). Single electrode deep brain stimulation with dual targeting Marano, M., Migliore, S., Squitieri, F., Insola, A., Scarnati, E., and Mazzone, P. at dual frequency for the treatment of chronic pain: a case series and review of (2019). CM-Pf deep brain stimulation and the long term management of motor the literature. Brain Sci. 7:E9. doi: 10.3390/brainsci7010009 and psychiatric symptoms in a case of Tourette syndrome. J. Clin. Neurosci. 62, Houeto, J. L., Karachi, C., Mallet, L., Pillon, B., Yelnik, J., Mesnage, V., et al. 269–272. doi: 10.1016/j.jocn.2018.12.029 (2005). Tourette’s syndrome and deep brain stimulation. J. Neurol. Neurosurg. Mazzone, P., Stocchi, F., Galati, S., Insola, A., Altibrandi, M. G., Modugno, N., et al. Psychiatry 76, 992–995. doi: 10.1136/jnnp.2004.043273 (2006). Bilateral implantation of centromedian-parafascicularis complex and

Frontiers in Human Neuroscience| www.frontiersin.org 6 January 2020 | Volume 13 | Article 447 Li et al. Imaging the Centromedian Nucleus

gpi: a new combination of unconventional targets for deep brain stimulation Velasco, F., Velasco, A. L., Velasco, M., Jimenez, F., Carrillo-Ruiz, J. D., and in severe Parkinson disease. 9, 221–228. doi: 10.1111/j.1525- Castro, G. (2007). Deep brain stimulation for treatment of the epilepsies: 1403.2006.00063.x the centromedian thalamic target. Acta Neurochir. Suppl. 97, 337–342. Morris, C. M., Candy, J. M., Oakley, A. E., Bloxham, C. A., and Edwardson, J. A. doi: 10.1007/978-3-211-33081-4_38 (1992). Histochemical distribution of non-haem iron in the . Acta Wang, Y., and Liu, T. (2015). Quantitative susceptibility mapping (QSM): Anat. 144, 235–257. doi: 10.1159/000147312 decoding MRI data for a tissue magnetic biomarker. Magn. Reson. Med. 73, Peppe, A., Gasbarra, A., Stefani, A., Chiavalon, C., Pierantozzi, M., Fermi, E., et al. 82–101. doi: 10.1002/mrm.25358 (2008). Deep brain stimulation of CM/PF of thalamus could be the new elective Wang, Y., Spincemaille, P., Liu, Z., Dimov, A., Deh, K., Li, J., et al. (2017). Clinical target for tremor in advanced Parkinson’s disease? Parkinsonism Relat. Disord. quantitative susceptibility mapping (QSM): biometal imaging and its emerging 14, 501–504. doi: 10.1016/j.parkreldis.2007.11.005 roles in patient care. J. Magn. Reson. Imaging 46, 951–971. doi: 10.1002/jmri. Savica, R., Stead, M., Mack, K. J., Lee, K. H., and Klassen, B. T. (2012). Deep 25693 brain stimulation in tourette syndrome: a description of 3 patients with Wei, H., Zhang, C., Wang, T., He, N., Li, D., Zhang, Y., et al. (2019). Precise excellent outcome. Mayo Clin. Proc. 87, 59–62. doi: 10.1016/j.mayocp.2011. targeting of the globus pallidus internus with quantitative susceptibility 08.005 mapping for deep brain stimulation surgery. J. Neurosurg. 11, 1–7. Schaltenbrand, G., Wahren, W., and Hassler, R. (1977). Atlas for Stereotaxy of the doi: 10.3171/2019.7.jns191254 Human Brain. Stuttgart: Thieme. Wei, H., Dibb, R., Zhou, Y., Sun, Y., Xu, J., Wang, N., et al. (2015). Streaking Sharma, V. D., Mewes, K., Wichmann, T., Buetefisch, C., Willie, J. T., and artifact reduction for quantitative susceptibility mapping of sources with large DeLong, M. (2017). Deep brain stimulation of the centromedian thalamic dynamic range. NMR Biomed. 28, 1294–1303. doi: 10.1002/nbm.3383 nucleus for essential tremor: a case report. Acta Neurochir. 159, 789–793. Wei, H., Cao, S., Zhang, Y., Guan, X., Yan, F., Yeom, K. W., et al. (2019). doi: 10.1007/s00701-017-3143-y Learning-based single-step quantitative susceptibility mapping reconstruction Son, B. C., Shon, Y. M., Choi, J. G., Kim, J., Ha, S. W., Kim, S. H., et al. without brain extraction. NeuroImage 202:116064. doi: 10.1016/j.neuroimage. (2016). Clinical outcome of patients with deep brain stimulation of the 2019.116064 centromedian thalamic nucleus for refractory epilepsy and location of the Wei, H., Zhang, Y., Gibbs, E., Chen, N. K., Wang, N., and Liu, C. (2017). Joint 2D active contacts. Stereotact. Funct. Neurosurg. 94, 187–197. doi: 10.1159/0004 and 3D phase processing for quantitative susceptibility mapping: application to 46611 2D echo-planar imaging. NMR Biomed. 30:e3501. doi: 10.1002/nbm.3501 Stefani, A., Peppe, A., Pierantozzi, M., Galati, S., Moschella, V., Stanzione, P., et al. Young, R. F., Vermeulen, S. S., Grimm, P., Posewitz, A. E., Jacques, D. B., (2009). Multi-target strategy for Parkinsonian patients: the role of deep brain Rand, R. W., et al. (1995). Gamma knife thalamotomy for the stimulation in the centromedian-parafascicularis complex. Brain Res. Bull. 78, treatment of persistent pain. Stereotact. Funct. Neurosurg. 64, 172–181. 113–118. doi: 10.1016/j.brainresbull.2008.08.007 doi: 10.1159/000098777 Testini, P., Zhao, C. Z., Stead, M., Duffy, P. S., Klassen, B. T., and Lee, K. H. (2016). Zhang, Y., Wei, H., Cronin, M. J., He, N., Yan, F., and Liu, C. (2018). Longitudinal Centromedian-parafascicular complex deep brain stimulation for tourette atlas for normative human brain development and aging over the lifespan using syndrome: a retrospective study. Mayo Clin. Proc. 91, 218–225. doi: 10.1016/j. quantitative susceptibility mapping. NeuroImage 171, 176–189. doi: 10.1016/j. mayocp.2015.11.016 neuroimage.2018.01.008 Testini, P., Min, H. K., Bashir, A., and Lee, K. H. (2016). Deep brain stimulation for tourette’s syndrome: the case for targeting the thalamic centromedian- Conflict of Interest: The authors declare that the research was conducted in the parafascicular complex. Front. Neurol. 7:193. doi: 10.3389/fneur.2016. absence of any commercial or financial relationships that could be construed as a 00193 potential conflict of interest. Tonge, M., Kocabicak, E., Ackermans, L., Kuijf, M., and Temel, Y. (2016). Final electrode position in subthalamic nucleus deep brain stimulation surgery: a Copyright © 2020 Li, Li, Gutierrez, Xu, Wu, Liu, Li, Sun, Zhang and Wei. This is an comparison of indirect and direct targeting methods. Turk. Neurosurg. 26, open-access article distributed under the terms of the Creative Commons Attribution 900–903. doi: 10.5137/1019-5149.JTN.13739-14.1 License (CC BY). The use, distribution or reproduction in other forums is permitted, Valentín, A., Garcia Navarrete, E., Chelvarajah, R., Torres, C., Navas, M., Vico, L., provided the original author(s) and the copyright owner(s) are credited and that the et al. (2013). Deep brain stimulation of the centromedian thalamic nucleus for original publication in this journal is cited, in accordance with accepted academic the treatment of generalized and frontal epilepsies. Epilepsia 54, 1823–1833. practice. No use, distribution or reproduction is permitted which does not comply doi: 10.1111/epi.12352 with these terms.

Frontiers in Human Neuroscience| www.frontiersin.org 7 January 2020 | Volume 13 | Article 447